Madon, S. (1997). What do people believe about gay males? A study of stereotype content and strength. Sex Roles, 37, 663-685.

What do people believe about gay males? A study of content and strength

Stephanie Madon

Rutgers University

This paper was partially funded by a grant from the Society for the Psychological Study of Social Issues (SPSSI) for dissertation research. Portions of this research were presented at the 67th Annual Conference of the Eastern Psychological Association, March 29 - 31, Philadelphia, PA. Special thanks are given to Max Guyll, Richard Ashmore, Alison Smith and David Wilder for their valuable comments on earlier versions of this article and to Shelly Keiper for her assistance in collecting the data. Correspondence regarding this article should be sent to Stephanie Madon, Department of Psychology, Tillett Hall, Room 101, Livingston Campus, Rutgers University, New Brunswick, NJ 08903.

Abstract

A multicomponent approach to stereotype assessment examined the content and strength of the gay male stereotype. There were 115 subjects in the main study (73 females and 42 males) including 9 African Americans, 19 Asians, 71 Caucasians, 10 Latinos and 6 subjects who did not report their ethnicity. Ninety-three subjects reported being heterosexual, 16 subjects reported that they were gay to some extent and 6 subjects did not report their sexual orientation. Analyses examining content assessed (1) stereotypic and counterstereotypic attributes of gay males including personality traits, behaviors and physical characteristics and; (2) different subtypes of gay males. Analyses examining strength measured how strongly people associated with gay males (1) personality traits, behaviors and physical characteristics and; (2) the identified subtypes. Results indicated that the stereotype’s content included attributes from multiple components and formed two subtypes. The first subtype reflected the perception that gay males have positive female sex-typed qualities. The second subtype reflected the perception that gay males violate acceptable male gender roles. Results regarding the stereotype’s strength indicated that people most strongly associated behaviors with gay males and the subtype that they violate acceptable male gender roles.

Psychology’s fascination with stereotypes can be traced back to the early part of this century. Walter Lippmann (1922) spoke of stereotypes as "pictures in our heads" and proposed that they have an important role in cognitive and perceptual processing. About a decade following Lippmann’s analysis, social psychologists began to empirically examine people’s stereotypes. The most renowned of these studies remains that conducted by Katz and Braly (1933). Princeton University students were presented with a list of attributes from which they selected five that most strongly characterized ten different ethnic and national groups. Katz and Braly’s research addressed two important questions: What do people believe about different social groups? And, how strongly do they believe it?

Although researchers have continued to address these questions since Katz and Braly (1933) first raised them (Gilbert, 1951; Karlins, Coffman, & Walters, 1969; see Brigham, 1971 for a review), interest in the content of stereotypes gradually declined and interest in the strength of stereotypes never really took off. The emphasis on content was replaced by an emphasis on process. Instead of asking "What attributes make up people’s stereotypes?" researchers began to ask such questions as "How do stereotypes bias perceptions?" (for reviews see Brewer, 1988; Fiske & Neuberg, 1990; Hamilton & Sherman, 1995; Stangor & Lange, 1994) and "What are the consequences of stereotyping?" (for a review see APA Brief, 1991). This emphasis has led to some important findings. Stereotypes may bias perceivers’ impressions of individuals (e.g., for reviews see, Brewer, 1988; Fiske & Neuberg, 1990), produce self-fulfilling prophecies (e.g., Snyder, Tanke & Bersheid, 1977; Word, Zanna, & Cooper, 1974) and lead to discrimination (APA Brief, 1991).

More recently, however, interest in the content of stereotypes has begun to re-emerge in the literature. There has been a resurgence of interest in the degree to which the content of stereotypes are accurate (Jussim, Eccles, & Madon, 1996), how the content of stereotypes change over time (Devine & Elliot, 1995), and in improving traditional techniques to assess the content of stereotypes (Stangor & Lange, 1994). The current research adds to this growing body of literature by assessing the content and strength of the gay male stereotype.

Why Study Stereotype Content And Strength?

Social psychology has long been concerned about the role that stereotypes play in social perception because of their potential to create social problems. It is this concern that has fueled the emphasis on issues of process. However, issues of content and process are complementary lines of inquiry. Stereotype content studies assess the specific attributes within stereotypes, their valence, strength and inaccuracy, and thus lay the necessary groundwork to examine the processes by which stereotypes may create social problems.

For example, determining the specific attributes within stereotypes makes it possible to measure stereotype change, an occurrence that is often regarded as a necessary step toward the reduction of prejudice (Allport, 1954). It also permits researchers to measure the valence of stereotypic attributes. This is important because not all stereotypes are negative. Consider the stereotypes that Asians are intelligent (e.g., Gilbert & Hixon, 1991; Katz & Braly, 1933), women are kind (Ashmore, Del Boca, & Wholers, 1986), and gay males are creative (Jackson & Sullivan, 1989). Predominantly negative stereotypes should have more damaging consequences on person perception than do predominantly positive stereotypes. Moreover, even predominantly positive stereotypes may negatively bias impressions if their negative components are held more strongly than their positive components. Content studies also make it possible to measure the inaccuracy of stereotypes, an issue that is crucial to establishing their self-fulfilling nature (Jussim, McCauley, & Lee, 1995). Perhaps it is not surprising, therefore, that researchers have been calling for more studies addressing issues of content (Stangor & Lange, 1994; Zebrowitz, 1996) and that such research is on the rise (e.g., Lee, Jussim, & McCauley, 1995).

What are Stereotypes?

Stereotypes are beliefs about the attributes that characterize a group of people (Ashmore & Del Boca, 1981). Stereotypes can be personal or consensual (Ashmore & Del Boca, 1981; Devine & Elliot, 1995). Personal stereotypes are one person’s beliefs about the attributes that characterize a group of people (e.g., Margaret believes that professors are absent minded). Consensual stereotypes are shared beliefs about the attributes that characterize a group of people (e.g., others agree with Margaret that professors are absent minded). Although the analyses reported herein focus exclusively on consensual stereotypes, the stereotype assessment technique presented is useful for measuring both personal and consensual stereotypes.

Stereotype content studies have traditionally emphasized personality traits to the exclusion of other stereotype components (e.g., with respect to gay males see Simmons, 1965; Staats, 1978: Stangor, Sullivan, & Ford, 1991; in general see, Brigham 1971 for a review). However, research indicates that stereotypes are multidimensional in nature, including several different components (Ashmore et al., 1986). For example, in addition to personality traits, stereotypes about gay males include physical characteristics, behaviors and attributes related to sexuality (Deaux & Lewis, 1984). A multicomponent approach to stereotype assessment, therefore, is necessary to fully assess the content of a stereotype.

A stereotype’s strength is conceptually distinct from its content. The content of a stereotype refers to the attributes that people believe characterize a group of people (Ashmore & Del Boca, 1981). The strength of a stereotype is the degree to which people believe the content of a stereotype (or its components) characterizes a group of people. That is, a stereotype’s strength refers to how strongly people associate a stereotype (or its components) with a social category (Stangor & Lange, 1994).

For example, imagine that people perceive Smithtown residents as very upper class, very snobbish and very reclusive, whereas they perceive Jonestown residents as slightly upper class, slightly snobbish and slightly reclusive. The contents of the stereotypes are the same. Both groups are perceived as upper class, snobbish and reclusive. However, the Smithtown stereotype is stronger than the Jonestown stereotype. People perceive the attributes in the Smithtown stereotype to be very characteristic of Smithtown residents, whereas they perceive the attributes in the Jonestown stereotype to be only slightly characteristic of Jonestown residents. Stereotypes, therefore, can be thought of in terms of content: The attributes that people believe characterize a group of people. Stereotypes can also be thought of in terms of strength: The degree to which people believe a set of attributes characterize a group of people.

Three procedures have been used to assess the content of the gay male stereotype: adjective checklists, rating scales, and free responses.

Adjective checklists. With adjective checklists people are presented with a predetermined list of attributes from which they select those that describe a social group’s members. Research that has used this procedure to assess the content of the gay male stereotype has found that people perceive gay males to be sexually abnormal, perverted, mentally ill, effeminate, lonely (Simmons, 1965), sensitive, individualistic, intelligent, honest, imaginative and neat (Staats, 1978). A strength of adjective checklists is that they can include a broad range of attributes, making them well suited to assessing several components within a stereotype. Adjective checklists have also been criticized, however, for having the potential to omit attributes that may be central people’s stereotypes (Stangor & Lange, 1994). For example, Staats (1978) assessed the gay male stereotype with an adjective checklist that was more than 40 years old and that was originally developed to assess the content of ethnic and national stereotypes. Results obtained from outdated or inappropriate adjective checklists may not accurately reflect the current stereotype content of a social group.

Rating scales. Rating scales are similar to adjective checklists in that people are also presented with a predetermined list of attributes. However, rather than indicating whether an attribute does, or does not, characterize a social group, people instead indicate the extent to which an attribute characterizes a social group. Studies that have used this procedure to assess the gay male stereotype have found that people perceive gay males to be compassionate, sensitive to the needs of others, creative, artistic (Jackson & Sullivan, 1989), mentally disturbed, prone to alcohol abuse (Jussim, Nelson, Manis & Soffin, 1995, Study 3), very feminine, likely to seduce young people (Herek, 1984; also see Herek, 1987), gentle, passive, theatrical (Gurwitz & Marcus, 1978) as having a strong need for security, liking art and literature (Page & Yee, 1986), having unusually strong sex drives, as being afraid of the opposite sex, and as acting like the opposite sex (Levitt & Klassen, 1974; for a review see Taylor, 1983).

Rating scales are similar to adjective checklists in that they can include a broad range of attributes, making them capable of assessing multiple components within a stereotype. They also have one distinct advantage over adjective checklists. Because subjects indicate the extent to which an attribute characterizes a social group, more precise measures of stereotypes (Ashmore et al., 1986), and in particular the strength of those stereotypes, can be obtained. However, rating scales also have the potential to omit attributes that may be central to people’s stereotypes (Stangor & Lange, 1994), a limitation that they share with adjective checklists.

Free responses. Free response procedures have people list attributes that they believe characterize a social group. Several studies have used this procedure to assess the content of the gay male stereotype and have found that people perceive gay males to be effeminate (Haddock, Zanna & Esses, 1993), as having a high pitched voice (Kite & Deaux, 1987), loving, abnormal, confused, and fussy (Stangor et al., 1991). The strength of this procedure is that it captures attributes that are central to people’s stereotypes. That is, people may list those attributes that they most strongly associate with a social group, and omit those attributes that do not readily come to mind. Unfortunately, however, this may lead to incomplete responding. People may not recall or record all of the attributes that they associate with a social group. As a result, this procedure may not fully assess people’s stereotypes.

Combined procedure. One procedure that has yet to be used to assess the content of the gay male stereotype is a combination of the adjective checklist, rating scale and free response procedures. Stangor and Lange (1994) suggest that such a combined procedure maximizes the strength of each individual procedure, while at the same time minimizing their weaknesses. Specifically, a free response procedure can identify the attributes that are most central to a stereotype. These free responses can then be combined with a larger list of attributes that include several different stereotype components. People can then indicate the extent to which each attribute on the combined list characterizes a social group. This combined procedure has been used to assess the content of sex stereotypes (Rosenkrantz, Vogel, Bee, Broverman & Broverman, 1968). The current study is the first to use this combined approach to assess the gay male stereotype.

Despite significant gains in understanding people’s beliefs about gay males, several additional issues deserve more attention. First, previous research has emphasized personality traits while neglecting other stereotype components (e.g., Page & Yee, 1986; Staats, 1978; Stangor et al., 1991; for exceptions see Kite & Deaux, 1987; Haddock et al., 1993). This is unfortunate because recent research suggests that behaviors and physical characteristics may be more central to stereotypes than are personality traits (McCauley, Jussim, & Lee, 1995). Second, although past research has identified a collection of stereotypic attributes, none have examined whether these attributes form underlying subtypes of gay males. Third, no research has identified attributes that are counterstereotypic of gay males. Fourth, although research to date clearly indicates that people have stereotypes about gay males, the strength of the stereotype has not been assessed.

This article addresses each of these issues. The content of the gay male stereotype was assessed by identifying those attributes that people perceived as stereotypic and counterstereotypic of gay males and by identifying different subtypes of gay males. The procedures that were used improved upon traditional stereotype assessment techniques by including attributes from multiple components (i.e. personality traits, behaviors and physical characteristics) and by combining all three stereotype assessment procedures (i.e. adjective checklist, rating scale and free response). Strength was examined by measuring how strongly people associated each of the stereotype components and the subtypes with gay males.

Preliminary Study One

Preliminary Study One was conducted as part of a larger study that used a free response procedure to assess the stereotypes of a variety of different social groups, including gay males.

Method

Ninety-eight (59 female, 37 male, 2 gender not reported) Rutgers University students took part in this study as one way of fulfilling a course requirement. There were 11 African Americans/Blacks, 22 Asians, 48 Whites, 11 other, and 6 subjects who did not report their ethnicity.

Subjects were run in large groups. After providing their informed consent, subjects completed a questionnaire that included the labels of 32 different social groups, presented in random order, of which "homosexuals" was one. For each group, subjects listed three attributes that they believed characterized the group’s members. Upon completion of the questionnaire, subjects were debriefed and dismissed.

Results

Five judges coded the free responses for two groups included in the questionnaire (i.e. homosexuals and Republicans). There were a total of 314 free responses (131 of which related to homosexuals) that were reduced in number with the following procedures. First, attributes listed only once were discarded to exclude idiosyncratic beliefs. Second, attributes that overlapped with Gough and Heilbrun’s (1983) adjective checklist were discarded because the entire checklist was included in Preliminary Study Two. Third, five judges met, and by consensus, discarded attributes that reflected social injustices (e.g., "are persecuted"). This was done because social injustices reflect perceived consequences of group membership rather than group members’ personal attributes. Fourth, the judges used group discussion to determine which of the remaining attributes were synonyms and then grouped the synonyms into categories. When consensus about the group in which an attribute should be categorized could not be reached, the attribute was left ungrouped. Finally, the judges used a majority decision rule to select one attribute from each group to reflect the group’s meaning. All other attributes in the groups were discarded. These procedures yielded 78 free responses that were combined with a predetermined list of attributes for use in Preliminary Study Two.

Preliminary Study Two

Preliminary Study Two included two parts. Part I identified whether people perceived a series of attributes as behaviors, physical characteristics, or neither. Part II used a rating scale procedure to identify attributes that people perceived to be the most stereotypic and the most counterstereotypic of gay males.

Method

Rutgers University students took part in this study as one way to fulfill a course requirement. Forty subjects (28 females and 12 males) participated in Part I, including 3 African Americans/Blacks, 3 Asians, 27 Whites, 6 other and 1 whose ethnicity was not reported. Of these subjects, 20 also participated in Part II (14 females and 6 males), including 15 Whites and 5 other. The remaining subjects participated in an unrelated activity.

The predetermined list consisted of 491 attributes, including the 78 free responses from Preliminary Study One, 300 personality traits from Gough and Heilbrun’s (1983) adjective checklist, and 113 experimenter-generated attributes. The experimenter-generated attributes were obtained from a systematic A through Z search of a pocket thesaurus (Roget’s II The New Thesaurus, Expanded Edition, 1988). Words that related to behaviors or physical characteristics were selected without regard to whether they described gay males specifically. This process yielded a pool of 151 attributes that was reduced to 113 by discarding synonyms, obscure words, and words that overlapped with subjects’ free responses or Gough and Heilbrun’s adjective checklist.

Part I. Subjects were run in pairs. After providing their informed consent, subjects completed a questionnaire that listed the 78 free responses and the 113 experimenter-generated attributes. Subjects were instructed to categorize each attribute as either a behavior, physical characteristic or neither.

Part II. Each of the 491 attributes was typed onto an index card and categorized by subjects into one of five containers labeled as "very uncharacteristic of male homosexuals," "somewhat uncharacteristic of male homosexuals," "no more characteristic of male homosexuals than of any other group," "somewhat characteristic of male homosexuals," and "very characteristic of male homosexuals." The order of the index cards was randomized prior to the start of each session. The tasks in Part I and Part II were counterbalanced for subjects who participated in both parts.

Results

Chi square analyses identified whether subjects perceived the free responses and experimenter-generated attributes as behaviors, physical characteristics, or neither. A significant Chi Square was reached if at least 27 out of 40 subjects assigned an attribute to the same category

Main Study

The main study addressed four issues. First, it assessed stereotypic and counterstereotypic attributes of gay males, including personality traits, behaviors, and physical characteristics. Second, it identified subtypes of gay males. Third, it compared the strength with which subjects associated personality traits, behaviors and physical characteristics with gay males. Fourth, it compared the strength with which subjects associated different subtypes with gay males.

Method

Subjects were 115 Rutgers University students (73 females and 42 males) who took part in this study as one way to fulfill a course requirement. There were 9 African Americans, 19 Asians, 71 Caucasians, 10 Latinos, and 6 subjects who did not report their ethnicity. There were 93 subjects who reported being "not at all homosexual," 1 who reported being "very much a homosexual," 15 who reported being homosexual to some extent and 6 who did not indicate their sexual orientation.

Subjects were run in large groups. After providing their informed consent, subjects completed a questionnaire that assessed, in the following order, demographic information (i.e., gender and ethnicity), their beliefs about male homosexuals, and their sexual orientation. Upon completion of the questionnaire, subjects were debriefed and dismissed.

The questionnaire included the 50 stereotypic and 25 counterstereotypic attributes identified in Preliminary Study Two. Subjects rated how characteristic they believed each attribute was of male homosexuals according to a five-point rating scale (1=very uncharacteristic of male homosexuals; 2=somewhat uncharacteristic of male homosexuals; 3=no more characteristic of male homosexuals than of any other group; 4=somewhat characteristic of male homosexuals; 5=very characteristic of male homosexuals).

Results

Analyses related to stereotype content addressed two questions: What do people believe about gay males? What subtypes of gay males do people hold?

Both factors consist of attributes related to the female gender role. However, the attributes loading on the first factor reflect positive qualities that do not necessarily violate acceptable male gender roles (e.g., warm-hearted, sensitive, good listeners). In sharp contrast, the attributes loading on the second factor reflect qualities that strongly violate acceptable male gender roles (e.g., limp wristed, transvestites, soft-voices, dainty). Thus, the results from the factor analysis suggest that gay males are perceived to exhibit positive female sex-typed qualities and to violate acceptable male gender roles.

Analyses addressed two questions related to stereotype strength: Do people more strongly associate personality traits, behaviors, or physical characteristics with gay males? Do people more strongly associate with gay males the subtype that they have positive female sex-typed qualities or that they violate acceptable male gender roles?

Calculating strength. The strength of the personality, behavior, and physical characteristic components were calculated by averaging how characteristic each subject perceived the set of 12 personality traits, the set of 6 behaviors, and the set of 6 physical characteristics to be of gay males. This yielded three means for each subject which will be referred to subsequently as the component strengths. The strength of the subtypes were calculated by averaging how characteristic each subject perceived the set of 18 attributes reflecting the perception that gay males have positive female sex-typed qualities and the set of 17 attributes reflecting the perception that gay males violate acceptable male gender roles. This yielded two means for each subject which will be referred to subsequently as the subtype strengths.

Between-subject differences. Repeated measures analysis of variance procedures (RM-ANOVAs) examined whether female versus male subjects and heterosexual versus gay subjects differed with regard to how strongly they associated the stereotype components and the subtypes with gay males. The between subjects factor was either gender (male, female) or sexual orientation (heterosexual, gay). The within factor was either component strengths (personality, behavior, physical characteristic) or subtype strengths (positive female sex-typed qualities, violation of acceptable male gender roles). In no case did the between subjects factors produce significant main effects (ps > .25), nor did they significantly interact with the within-subjects factors (ps > .14). Therefore, all subsequent analyses collapsed across gender and sexual orientation.

Do people more strongly associate with gay males the subtype that they have positive female sex-typed qualities or the subtype that they violate acceptable male gender roles? A dependent samples t-test examined whether the two subtypes differed in strength. The dependent variables were the subtype strengths. Results indicated that the subtype that gay males violate acceptable male gender roles (M=4.03) was significantly stronger than the subtype that gay males have positive female sex-typed qualities (M=3.93; t₍₁₀₆₎ = 2.18; p = .03).

Discussion

What do people believe about gay males, and how strongly do they believe it? This study answered these questions by assessing the content and strength of the gay male stereotype. Results regarding the content of the stereotype found that people's beliefs about gay males include a broad and rich set of attributes that form two general subtypes -- that gay males have positive female sex-typed qualities and that they violate acceptable male gender roles. Results regarding the strength of the stereotype found that people most strongly associated behaviors with gay males and the belief that they violate acceptable male gender roles. The relation of these findings to past research and their implications for biases in social perception are discussed.

Stereotypic and counterstereotypic attributes. The current study found results consistent with past research on the content of the gay male stereotype. For example, gay males were perceived as talkative, gentle, fashionable and artistic, a result that is very similar to that reported by Page and Yee (1986) who found that people judged gay males as talkative, gentle, concerned about their appearance and liking art and literature. Both the current study and Staats (1978) found that people judged gay males as sensitive and individualistic. Finally, the current study found that people judged gay males as liberal and different, a finding that is comparable to the results of Stangor et al. (1991) which indicated that people judged gay males as liberal and abnormal.

The current study also found, however, that the gay male stereotype included attributes that have not previously been identified. Many of these attributes were behaviors and physical characteristics. For example, gay males were perceived to engage in anal sex, be open about their feelings, melodramatic, artsy looking, and to wear earrings. Other attributes were counterstereotypic of gay males. For example gay males were perceived to NOT act macho, pick fights, or hunt animals.

The identification of these previously unidentified attributes reflects the use of an improved stereotype assessment procedure. The gay male stereotype was assessed with a combination of adjective checklists, rating scales and free responses that included attributes from multiple stereotype components. This resulted in a broader and richer taxonomy of beliefs about gay males than has previously been found. The identification of these additional attributes is important from a methodological standpoint because the more researchers know about what people believe, the more places that they may detect bias. In turn, detecting where bias does and does not occur has theoretical implications because it offers insights into when biases are most probable and when their effects may be most damaging.

Subtypes of gay males. Past research on the content of the gay male stereotype has identified a collection of stereotypic attributes. The current research added to this literature by examining whether attributes stereotypic of gay males formed different subtypes. Results indicated the presence of two subtypes, both of which included female sex-typed attributes. This finding is consistent with past research which shows that people perceive gay males to be like woman, referred to as inversion theory (Kite & Deaux, 1987). However, the two subtypes differed from each other in an important respect. Whereas the first reflected the perception that gay males exhibit positive female sex-typed qualities, the second reflected the perception that gay males exhibit female sex-typed qualities that violate acceptable male gender roles. The positive and negative valence of these subtypes suggest that bias against gay males might arise more from the negative perception that gay males violate what it means to be a man than from the positive perception that gay males possess favorable qualities associated with women.

This study addressed the strength of the gay male stereotype by examining how strongly people associated three stereotype components and two subtypes with gay males. Results indicated that people associated behaviors with gay males more strongly than they associated personality traits and physical characteristics with gay males. People also associated the subtype that gay males violate acceptable male gender roles more strongly than the subtype that they have positive female sex-typed qualities. However, because the magnitude of these differences was small their practical effects should be interpreted with caution. Nonetheless, information about a stereotype's strength is important for several reasons.

Strongly held stereotypes may bias person perception more than weakly held stereotypes. This is because people may be more confident that strongly held beliefs accurately reflect a group’s actual attributes, increasing the likelihood that they will make stereotypic inferences. Additionally, people may perceive group members as relatively homogeneous on attributes that they strongly associate with a group. Indeed, recent research indicates that the more homogeneous a group’s members are perceived to be on a particular attribute, the more perceivers rely on stereotypes during person perception (Ryan, Judd & Park, 1996). Strongly held stereotypes may also increase the power of self-fulfilling prophecies because people may be less motivated to revise beliefs that they hold strongly (Jussim et al, 1996).

There are several important issues concerning the gay male stereotype that future research may want to explore. First, research that identifies the valence of the specific attributes in the gay male stereotype is needed because only negatively biased attributes should produce unfavorable perceptions. This is particularly important in light of research suggesting that perceivers' affective reactions to gay males influence stereotyping more than do their cognitive beliefs (Jussim et al., 1995, Study 3). A second issue that requires more attention is the inaccuracy of the gay male stereotype. Although often not tackled because of its methodological complexity (Jussim et al., 1995), research on stereotype inaccuracy is very important since only inaccurate stereotypes can create biases (Jussim et al, 1996). Third, the vast majority of research on stereotyping, including the current studies, use college students as subjects. However, college students may differ from other populations in important respects because of their age, maturity, education, socio-economic status, levels of liberalism, and prejudice toward gay males, etc. Of particular importance, therefore, is research that examines the extent to which results based on college samples generalizes to other populations.

Conclusion

The current research examined the content and strength of the gay male stereotype. This focus represents a return to the earliest issues addressed by social psychologists. Although once a major substantive area in social psychology, issues of content gradually gave way to issues of process. However, issues of content and process are inextricably tied to one another. Content studies detail the specific attributes in stereotypes, their strength, valence, and inaccuracy. Process studies use this information to examine when stereotypes will influence social reality and social perception. The current research showed that beliefs about gay males included attributes from multiple stereotype components, formed two subtypes and varied in strength. These findings have implications for research on issues of process because knowing what people believe about gay males and how strongly they hold those beliefs provides insight into when stereotypes may be most likely to create biases.

References

Allport, G. W. (1954). The nature or prejudice. Cambridge, MA: Addison Wesley.

APA Amicus Curiae Brief for the American Psychological Association (1991). In the Supreme Court of the United States: Price Waterhouse v. Ann B. Hopkins. American Psychologist, 46, 1061-1070.

Ashmore, R. D., & Del Boca, F. K. (1981). Conceptual approaches to stereotypes and stereotyping. In D. Hamilton (Ed.), Cognitive processes in stereotyping and intergroup behavior, (pp. 37-81). Erlbaum Publishers: Hillsdale, NJ.

Ashmore, R. D., & Del Boca, F. K., & Wohlers, A. J. (1986). Gender stereotypes. In R. D. Ashmore & F. K. Del Boca (Eds.), The social psychology of female-male relations: A critical analysis of central concepts (pp. 69-119). New York: Academic Press.

Brewer, M. B. (1988). A dual process model of impression formation. In T. K. Srull & R. S. Wyer, Jr., (Eds.), Advances in social cognition, (Vol. 1). Lawrence Erlbaum Associates, Publishers: Hillsdale, NJ.

Brigham, J. C. (1971). Ethnic stereotypes. Psychological Bulletin, 76, 15-33.

Coleman, E. (1987). Assessment of sexual orientation. Journal of Homosexuality, 14, 9-24.

Deaux, K. & Lewis, L. L. (1984). Structure of gender stereotypes: Interrelationships among components and gender label. Journal of Personality and Social Psychology, 46, 991-1004.

Devine, P. G., Elliot, A. J. (1995). Are racial stereotypes really fading? The Princeton trilogy revisited. Personality and Social Psychology Bulletin, 21, 1139-1150.

Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression formation, from category-based to individuating processes: Influences of information and motivation on attention and interpretation. In M. P. Zanna (Ed.), Advances in experimental social psychology, (Vol. 23, pp. 1-74). New York: Academic Press.

Gilbert, D. T., & Hixon, J. G. (1991). The trouble of thinking: Activation and application of stereotypic beliefs. Journal of Personality and Social

Psychology, 60, 509-517.

Gilbert, G. M. (1951). Stereotype persistence and change among college students. Journal of Abnormal and Social Psychology, 46, 245-254.

Gough, H. G., & Heilbrun, A. B. (1983). In O. K. Buros (Ed.), The Adjective Checklist Manual. Palo Alto, CA: Consulting Psychologists Press.

Gurwitz, S. B., & Marcus, M. (1978). Effects of anticipated interaction, sex and homosexual stereotypes on first impressions. Journal of Applied Social Psychology, 90, 173-183.

Haddock, G., Zanna, M. P., & Esses, V. M. (1993). Assessing the structure of prejudicial attitudes: The case of attitudes toward homosexuals. Journal of Personality and Social Psychology, 65, 1105-1118.

Hamilton, D. L., & Sherman, J. W. (1995). Stereotypes. In R. S. Wyer & T. K. Srull (Eds.), Handbook of social cognition, (2nd Edition, pp. 1-68). Hillsdale, New Jersey: Erlbaum.

Herek, G. M. (1984). Attitudes toward lesbians and gay men: A factor-analytic study. Journal of Homosexuality 10, 39-52.

Herek, G. M. (1987). Can functions be measured? A new perspective on the functional approach to attitudes. Social Psychology Quarterly, 50, 285-303.

Jackson, L. A., Sullivan, L. A. (1989). Cognition and affect in evaluations of stereotyped group members. The Journal of Social Psychology, 129, 659-672.

Jussim, L., Eccles, J., & Madon, S. (1996). Social perception, social stereotypes, and teacher expectations: Accuracy and the quest for the powerful self-fulfilling prophecy. In M. P. Zanna (Ed.), Advances in experimental social psychology, (Vol. 28, pp. 281-388). San Diego, California: Academic Press.

Jussim, L., Nelson, T., Manis, M., & Soffin, S. (1995). Prejudice, stereotypes, and labeling effects: Sources of bias in person perception. Journal of Personality and Social Psychology, 68, 228-246.

Katz, D. & Braly, K. W. (1933). Racial stereotypes of one hundred college students. Journal of Abnormal and Social Psychology, 28, 280-290.

Karlins, M., Coffman, T. L., & Walters, G. (1969). On the fading of social stereotypes: Studies in three generations of college students. Journal of Personality and Social Psychology, 13, 1-10.

Kite, M. E., & Deaux, K. (1987). Gender belief systems: Homosexuality and the implicit inversion theory. Psychology of Women Quarterly, 11, 83-96.

Lee, Y., Jussim, L. J., & McCauley, C. R. (1995). Stereotype accuracy: Toward appreciating group differences. Washington, D.C.: American Psychological Association.

Levitt, E. E., & Klassen, A. D., Jr. (1974). Public attitudes toward homosexuality: Part of the 1970 national survey by the Institute for Sex Research. Journal of Homosexuality, 1, 29-43.

Lippmann, W. (1922) Public opinion. New York: Harcourt Brace.

Masters, W. H., Johnson, V. E., & Kolodny, R. C. (1986). Masters and Johnson on sex and human loving. Boston, MA: Little, Brown and Company.

McCauley, C. R., Jussim, L. J., & Lee, Y. (1995). Stereotype accuracy: Toward appreciating group differences. In Y. Lee, L. J. Jussim and C. R. McCauley (Eds.), Stereotype accuracy: Toward appreciating group differences (pp. 293-312). Washington, D.C.: American Psychological Association.

Page, S., Yee, M. (1986). Conception of male and female homosexual stereotypes among university undergraduates. Journal of Homosexuality, 12, 109-117.

Roget’s II: The new thesaurus. (1988). New York, NY: Berkley Books.

Rosenkrantz, P., Vogel, S., Bee, H., Broverman, I., & Broverman, D. (1968). Sex-role stereotypes and self-concepts in college students. Journal of Consulting and Clinical Psychology, 32, 287-295.

Ryan, C. S., Judd, C. M., & Park, B. (1996). Effects of racial stereotypes on judgments of individuals: The moderating role of perceived group variability. Journal of Experimental and Social Psychology, 32, 71-103.

Simmons, J. L. (1965). Public stereotypes of deviants. Social Problems, 13, 223-232.

Snyder, M., Tanke, E. D., & Berscheid, E. (1977). Social perception and interpersonal behavior: On the self-fulfilling nature of social stereotypes. Journal of Personality and Social Psychology, 35, 656-666.

Staats, G. R. (1978). Stereotype content and social distance: Changing views of homosexuality. Journal of Homosexuality, 4, 15-28.

Stangor, C., & Lange, J. E. (1994). Mental representations of social groups: Advances in understanding stereotypes and stereotyping. In M. P. Zanna (Ed.), Advances in experimental social psychology, (Vol. 26, pp. 357-416). San Diego, California: Academic Press.

Stangor, C., Sullivan, L. A., & Ford, T. (1991). Affective and cognitive determinants of prejudice. Social Cognition, 9, 359-380.

Taylor, A. (1983). Conceptions of masculinity and femininity as a basis for stereotypes of male and female homosexuals. Journal of Homosexuality, 9, 37-53.

Word, C. O., Zanna, M. P., & Cooper, J. (1974). The nonverbal mediation of self-fulfilling prophecies in interracial interaction. Journal of Experimental Social Psychology, 10, 109-120.

Zebrowitz, L. A. (1996). Physical Appearance as a basis of stereotyping. In C. N. Macrae, C. Stangor, & M. Hewstone (Eds.), Stereotypes and stereotyping (pp. 79-120). New York, NY: Guilford Press.

To view Table 1:

Footnotes

1. The 32 social groups were presented in the following order: sports fanatics, bikers, nerds, fashion models, politicians, party animals, punk rockers, fraternity members, sorority members, conservatives, liberals, truck drivers, body builders, strippers, computer scientists, democrats, republicans, war veterans, animal rights activists, single parents, homosexuals, jocks, deadheads, ex-cons, secretaries, vegetarians, pro-life supporters, environmentalists, born again christians, skin heads, construction workers, people infected with HIV.

2. Preliminary Study One was part of a larger study that was designed to identify the stereotype content of two groups that subjects liked equally but with whom they had different levels of personal contact. Homosexuals and Republicans met this criteria. Thus, free responses for both groups were coded and included in Preliminary Study Two. Interestingly, the attributes perceived as stereotypic of Republicans were often perceived as counter-stereotypic of gay males.

3. Gough and Heilbrun’s (1983) adjective checklist contains 300 words and phrases that encompass a broad range of personality traits.

4. Sexual orientation was determined with a single item that asked subjects: "To what degree do you consider yourself a homosexual?" Responses were made on a five point scale with endpoints "not at all a homosexual" and "very much a homosexual." The use of a rating scale to measure sexual orientation is preferred over dichotomous and trichotomous measures (Coleman, 1987). However, there is also broad consensus that continuous measures of sexual orientation should distinguish between different aspects of homosexuality (e.g., behaviors versus sexual fantasies; see Coleman, 1987). Although the current study only measured sexual orientation on one dimension, results were nonetheless consistent with national statistics. Specifically, 14% of subjects in the current study and 13% of Americans report being homosexual to some extent (Masters, Johnson & Kolodny, 1986).

5. In the main study, "wear flashy clothes" did meet the 60% criteria with 65.2% of subjects judging it as very or somewhat characteristic of gay males. However, because 11.3% of subjects judged it as very or somewhat uncharacteristic of gay males, it did not meet the 10% criteria and, therefore, was not included as part of the gay male stereotype.

6. A five point questionnaire item (see footnote 4) measured sexual orientation. However, 93 out of 109 subjects indicated that they were not at all gay (response option = 1). Because of the small sample size at each of the remaining levels, subjects were categorized as gay if they reported being gay to some extent (n=16) and as heterosexual if they reported that they were not at all gay (n=93).

7. In contrast to the two subtypes that emerged here, Herek (1984) found that a single factor, "condemnation - tolerance," explained people's attitudes about gay males and lesbians. However, the "condemnation - tolerance" factor reflected attitude statements (e.g., "Male homosexuality is a perversion" pg. 51), whereas the subtypes (i.e. factors) that emerged in the current research reflected beliefs about gay males. Indeed, Herek himself found that people's beliefs about gay males and lesbians formed a second factor in addition to the "condemnation tolerance" factor. This strongly suggests that people's attitudes and beliefs about gay males and lesbians are distinct constructs.